Skip to content

Retarget cross-provider duplicate book rows to the grabbed book on shared edition identity #113

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jbob06 wants to merge 1 commit into
base: develop
Choose a base branch
from
Open

Retarget cross-provider duplicate book rows to the grabbed book on shared edition identity #113

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
237 changes: 237 additions & 0 deletions src/Chaptarr.Core.Test/MediaFiles/DownloadedBooksImportServiceFixture.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1826,6 +1826,243 @@ public void should_reject_same_work_sibling_match_when_grabbed_book_has_no_equiv
}
}

// Issue #111: an author-bibliography re-discovery re-added the same real book under a different
// metadata provider, so the duplicate row's work ID (gr:...) never intersects the grabbed row's
// work ID (hc:...). Both rows still carry the same physical edition (same Audible ASIN), which is
// globally unique, so the download must be retargeted onto the row that actually owns the grab.
[Test]
public void should_retarget_cross_provider_duplicate_row_match_to_grabbed_book_when_edition_identity_matches()
{
var tempDir = Path.Combine(Path.GetTempPath(), "chaptarr-tests", Guid.NewGuid().ToString("N"));
Directory.CreateDirectory(tempDir);
var filePath = Path.Combine(tempDir, "Fate Hollow Academy Term 1.m4b");
File.WriteAllBytes(filePath, new byte[] { 1, 2, 3, 4 });

try
{
var diskProvider = new StubDiskProvider();
var diskScanService = new StubDiskScanService();
var tagsService = new StubMetadataTagService();

var matchingService = new StubFileMatchingService
{
InitialResult = new FileMatchResult
{
MatchedFiles = new[]
{
new FileMatch
{
File = new DiscoveredFileWithMetadata { Path = filePath, Size = 4, Modified = DateTime.UtcNow, AllTags = tagsService.Tags },
AuthorId = 7,
AuthorName = "Test Author",
BookId = 29100,
BookTitle = "Fate Hollow Academy",
EditionId = 30
}
},
UnmatchedFiles = Array.Empty<UnmatchedFile>()
}
};

var author = new Author { Id = 7, Name = "Test Author" };

// Never-monitored duplicate row, re-discovered from Goodreads.
var matchedBook = new Book { Id = 29100, AuthorId = 7, Author = author, Title = "Fate Hollow Academy", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, GoodreadsWorkId = "gr:200474660" };

// The row that was actually grabbed, sourced from Hardcover. No work-ID overlap with the duplicate.
var targetBook = new Book { Id = 29083, AuthorId = 7, Author = author, Title = "Fate Hollow Academy: Term 1", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, HardcoverBookId = "hc:585413" };

var matchedEdition = new Edition { Id = 30, BookId = 29100, Title = "Fate Hollow Academy", AudibleASIN = "B0DXYZ1234", ReadingFormatId = 2, Monitored = false, Book = matchedBook };
var targetEdition = new Edition { Id = 40, BookId = 29083, Title = "Fate Hollow Academy: Term 1", AudibleASIN = "B0DXYZ1234", ReadingFormatId = 2, Monitored = false, Book = targetBook };
matchedBook.Editions = new List<Edition> { matchedEdition };
targetBook.Editions = new List<Edition> { targetEdition };

var importApproved = new RecordingImportApprovedBooks();

var bookService = DispatchProxy.Create<IBookService, BookServiceProxy>();
var bookProxy = (BookServiceProxy)(object)bookService;
bookProxy.BooksById[29100] = matchedBook;
bookProxy.BooksById[29083] = targetBook;

var authorService = DispatchProxy.Create<IAuthorService, AuthorServiceProxy>();
((AuthorServiceProxy)(object)authorService).Author = author;

var editionService = DispatchProxy.Create<IEditionService, EditionServiceProxy>();
var editionProxy = (EditionServiceProxy)(object)editionService;
editionProxy.EditionsById[30] = matchedEdition;
editionProxy.EditionsById[40] = targetEdition;
editionProxy.EditionsByBookId[29100] = new List<Edition> { matchedEdition };
editionProxy.EditionsByBookId[29083] = new List<Edition> { targetEdition };

var importOrchestrator = DispatchProxy.Create<IImportOrchestrator, ThrowingProxy<IImportOrchestrator>>();
var configService = ConfigServiceTestProxy.Create();
var historyService = DispatchProxy.Create<IHistoryService, HistoryServiceProxy>();
var eventAggregator = DispatchProxy.Create<IEventAggregator, ThrowingProxy<IEventAggregator>>();
var runtimeInfo = DispatchProxy.Create<NzbDrone.Common.EnvironmentInfo.IRuntimeInfo, ThrowingProxy<NzbDrone.Common.EnvironmentInfo.IRuntimeInfo>>();

var service = new DownloadedBooksImportService(
diskProvider,
diskScanService,
matchingService,
tagsService,
importApproved,
bookService,
authorService,
editionService,
importOrchestrator,
new StubAuthorLibraryService(),
new StubRootFolderService(),
configService,
historyService,
eventAggregator,
runtimeInfo,
DispatchProxy.Create<IMediaInfoExtractor, ThrowingProxy<IMediaInfoExtractor>>(),
LogManager.GetCurrentClassLogger());

var remoteBook = new RemoteBook
{
Author = author,
Books = new List<Book> { new Book { Id = 29083, AuthorId = 7, Title = "Fate Hollow Academy: Term 1", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, HardcoverBookId = "hc:585413" } }
};

var downloadClientItem = new DownloadClientItem
{
DownloadId = "DOWNLOAD-CROSS-PROVIDER-DUPLICATE",
CanMoveFiles = false,
DownloadClientInfo = new DownloadClientItemClientInfo { Id = 1, Name = "qBittorrent", Type = "qBittorrent" }
};

_ = service.ProcessPath(filePath, ImportMode.Auto, author, downloadClientItem, remoteBook);

Assert.That(importApproved.Decisions, Has.Count.EqualTo(1));
Assert.That(importApproved.Decisions[0].Approved, Is.True);
Assert.That(importApproved.Decisions[0].Item.Book.Id, Is.EqualTo(29083));
Assert.That(importApproved.Decisions[0].Item.Edition.Id, Is.EqualTo(40));
Assert.That(importApproved.Decisions[0].Item.Edition.Book, Is.SameAs(targetBook));
}
finally
{
try { Directory.Delete(tempDir, recursive: true); } catch { }
}
}

// Safety boundary for issue #111: same author and media type, but the two rows share no work-level
// ID and no edition-level ID. Nothing proves they are the same book, so the file must NOT be moved
// onto the grabbed row -- the download is rejected and the matched row is left untouched.
[Test]
public void should_not_retarget_cross_provider_row_match_when_no_shared_work_or_edition_identity()
{
var tempDir = Path.Combine(Path.GetTempPath(), "chaptarr-tests", Guid.NewGuid().ToString("N"));
Directory.CreateDirectory(tempDir);
var filePath = Path.Combine(tempDir, "Some Other Book.m4b");
File.WriteAllBytes(filePath, new byte[] { 1, 2, 3, 4 });

try
{
var diskProvider = new StubDiskProvider();
var diskScanService = new StubDiskScanService();
var tagsService = new StubMetadataTagService();

var matchingService = new StubFileMatchingService
{
InitialResult = new FileMatchResult
{
MatchedFiles = new[]
{
new FileMatch
{
File = new DiscoveredFileWithMetadata { Path = filePath, Size = 4, Modified = DateTime.UtcNow, AllTags = tagsService.Tags },
AuthorId = 7,
AuthorName = "Test Author",
BookId = 29100,
BookTitle = "A Different Book",
EditionId = 30
}
},
UnmatchedFiles = Array.Empty<UnmatchedFile>()
}
};

var author = new Author { Id = 7, Name = "Test Author" };
var matchedBook = new Book { Id = 29100, AuthorId = 7, Author = author, Title = "A Different Book", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, GoodreadsWorkId = "gr:200474660" };
var targetBook = new Book { Id = 29083, AuthorId = 7, Author = author, Title = "Fate Hollow Academy: Term 1", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, HardcoverBookId = "hc:585413" };

var matchedEdition = new Edition { Id = 30, BookId = 29100, Title = "A Different Book", AudibleASIN = "B0DAAAAAAA", ReadingFormatId = 2, Monitored = false, Book = matchedBook };
var targetEdition = new Edition { Id = 40, BookId = 29083, Title = "Fate Hollow Academy: Term 1", AudibleASIN = "B0DZZZZZZZ", ReadingFormatId = 2, Monitored = false, Book = targetBook };
matchedBook.Editions = new List<Edition> { matchedEdition };
targetBook.Editions = new List<Edition> { targetEdition };

var importApproved = new RecordingImportApprovedBooks();

var bookService = DispatchProxy.Create<IBookService, BookServiceProxy>();
var bookProxy = (BookServiceProxy)(object)bookService;
bookProxy.BooksById[29100] = matchedBook;
bookProxy.BooksById[29083] = targetBook;

var authorService = DispatchProxy.Create<IAuthorService, AuthorServiceProxy>();
((AuthorServiceProxy)(object)authorService).Author = author;

var editionService = DispatchProxy.Create<IEditionService, EditionServiceProxy>();
var editionProxy = (EditionServiceProxy)(object)editionService;
editionProxy.EditionsById[30] = matchedEdition;
editionProxy.EditionsById[40] = targetEdition;
editionProxy.EditionsByBookId[29100] = new List<Edition> { matchedEdition };
editionProxy.EditionsByBookId[29083] = new List<Edition> { targetEdition };

var importOrchestrator = DispatchProxy.Create<IImportOrchestrator, ThrowingProxy<IImportOrchestrator>>();
var configService = ConfigServiceTestProxy.Create();
var historyService = DispatchProxy.Create<IHistoryService, HistoryServiceProxy>();
var eventAggregator = DispatchProxy.Create<IEventAggregator, ThrowingProxy<IEventAggregator>>();
var runtimeInfo = DispatchProxy.Create<NzbDrone.Common.EnvironmentInfo.IRuntimeInfo, ThrowingProxy<NzbDrone.Common.EnvironmentInfo.IRuntimeInfo>>();

var service = new DownloadedBooksImportService(
diskProvider,
diskScanService,
matchingService,
tagsService,
importApproved,
bookService,
authorService,
editionService,
importOrchestrator,
new StubAuthorLibraryService(),
new StubRootFolderService(),
configService,
historyService,
eventAggregator,
runtimeInfo,
DispatchProxy.Create<IMediaInfoExtractor, ThrowingProxy<IMediaInfoExtractor>>(),
LogManager.GetCurrentClassLogger());

var remoteBook = new RemoteBook
{
Author = author,
Books = new List<Book> { new Book { Id = 29083, AuthorId = 7, Title = "Fate Hollow Academy: Term 1", AnyEditionOk = true, MediaType = BookMediaType.Audiobook, HardcoverBookId = "hc:585413" } }
};

var downloadClientItem = new DownloadClientItem
{
DownloadId = "DOWNLOAD-CROSS-PROVIDER-UNRELATED",
CanMoveFiles = false,
DownloadClientInfo = new DownloadClientItemClientInfo { Id = 1, Name = "qBittorrent", Type = "qBittorrent" }
};

_ = service.ProcessPath(filePath, ImportMode.Auto, author, downloadClientItem, remoteBook);

Assert.That(importApproved.Decisions, Has.Count.EqualTo(1));
Assert.That(importApproved.Decisions[0].Approved, Is.False);

// The file must stay attached to the row the matcher chose; retargeting it would misfile it.
Assert.That(importApproved.Decisions[0].Item.Book.Id, Is.EqualTo(29100));
Assert.That(importApproved.Decisions[0].Rejections.Select(r => r.Reason), Has.Some.Contains("but import matched"));
Assert.That(importApproved.Decisions[0].Rejections.Select(r => r.Reason), Has.None.Contains("no equivalent edition"));
}
finally
{
try { Directory.Delete(tempDir, recursive: true); } catch { }
}
}

[Test]
public void should_reject_completed_download_when_latest_grabbed_edition_is_strict_and_match_hits_different_sibling()
{
Expand Down
40 changes: 37 additions & 3 deletions src/NzbDrone.Core/MediaFiles/DownloadedBooksImportService.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -725,15 +725,44 @@ private void RetargetSameWorkMatchesToGrabbedBook(List<ImportDecision<LocalBook>
{
var localBook = decision.Item;
var matchedBook = localBook?.Book;
if (!CanRetargetSameWorkMatch(targetBook, matchedBook))
if (!SharesRetargetableRowIdentity(targetBook, matchedBook))
{
continue;
}

var matchedEdition = localBook.Edition;

// Two independent same-book signals are accepted here:
//
// 1. Work-level provider IDs intersect (hc/gr/ol work IDs). This is the original signal and
// covers duplicate rows catalogued through the same metadata provider.
//
// 2. The matched edition itself also exists under the grabbed book row, identified by an
// edition-level provider ID (ISBN-10/13, ASIN/Audible ASIN, or a Goodreads/Hardcover/
// OpenLibrary/Google Books edition ID). Duplicate rows created by author-bibliography
// re-discovery are frequently sourced from a *different* provider than the row that was
// grabbed, so their work-level ID sets never intersect even though they describe the same
// title. Signal 2 recognises those without weakening the guarantee: edition-level provider
// IDs are globally unique per physical edition, so a shared one means both rows carry the
// very same edition. Combined with the same-author and same-media-type guards in
// SharesRetargetableRowIdentity, this cannot collide two different books that merely share
// a title -- which is exactly why title similarity is deliberately NOT used as a signal.
//
// Both signals still have to land on a concrete equivalent edition below before anything is
// retargeted, so a match on either one can only ever move the file to an edition the grabbed
// book row already owns.
var sameWorkProviderId = WorkIdMatcher.WorkProviderIdMatches(targetBook, matchedBook);
var targetEdition = FindEquivalentEditionForTargetBook(targetBook, matchedEdition);

if (targetEdition == null)
{
if (!sameWorkProviderId)
{
// Neither signal fired: nothing establishes that these two rows are the same book.
// Leave the decision untouched so the caller's expected-book check rejects it.
continue;
}

decision.Reject(new Rejection(
$"Completed download was grabbed for {FormatBookLabel(targetBook)} and matched same-work sibling {FormatBookLabel(matchedBook)}, but no equivalent edition exists under the grabbed book row. Refresh metadata and retry."));
continue;
Expand Down Expand Up @@ -780,7 +809,12 @@ private Book HydrateExpectedBook(Book book)
}
}

private static bool CanRetargetSameWorkMatch(Book targetBook, Book matchedBook)
/// <summary>
/// Baseline guards every retarget must clear, independent of how "same book" is established:
/// two distinct persisted rows, the same media type, and the same (known) author. These are
/// hard requirements -- retargeting across authors or across media types would misfile the download.
/// </summary>
private static bool SharesRetargetableRowIdentity(Book targetBook, Book matchedBook)
{
if (targetBook == null ||
matchedBook == null ||
Expand All @@ -803,7 +837,7 @@ private static bool CanRetargetSameWorkMatch(Book targetBook, Book matchedBook)
return false;
}

return WorkIdMatcher.WorkProviderIdMatches(targetBook, matchedBook);
return true;
}

private Edition FindEquivalentEditionForTargetBook(Book targetBook, Edition matchedEdition)
Expand Down